Date: Wed, 04 Sep 1996 10:13:55 -0700 From: John Polstra <jdp@polstra.com> To: jkh@time.cdrom.com Cc: current@FreeBSD.org, sja@tekla.fi Subject: Re: Latest Current build failure Message-ID: <199609041713.KAA18153@austin.polstra.com> In-Reply-To: <199609040346.UAA14084@austin.polstra.com>
next in thread | previous in thread | raw e-mail | index | archive | help
I was thinking some more about an idea Jordan brought up: > Actually, all you'd need to do is have your cvsup file modified to > use a specific date in updating a previously checked out version > of HEAD. Here is a related idea I've been toying with. I think I could easily extend CVSup to support "pseudo-tags". These would be like CVS tags, except they wouldn't really exist in the repository. Instead, the CVSup server would know about them. A pseudo-tag would correspond to some date on some branch. The user could request an update for a pseudo-tag exactly as he now gets an update for a real tag. There could be a pseudo-tag called, say, RELENG_2_2_RECENT, representing the almost-current, more-likely-buildable version. I.e., it would refer to the main branch at some cutoff date specified on the server. The cutoff date could be read from a configuration file on the server. Whenever a newer version of -current was deemed to be good/stable enough, the date would be updated in the server configuration file. The user would always specify the same tag; it would mean whatever the people who administered the server wanted it to mean. That seems better than asking the users to change their cutoff dates all the time. John -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199609041713.KAA18153>